Background: Intraventricular hemorrhage (IVH) is a major cause of neurologic disabilities in preterm newborns. We evaluated the use of plasma activin A concentrations to predict the development of perinatal IVH.
Methods: We measured nucleated erythrocyte (NRBC) counts, plasma activin A, hypoxanthine (Hyp), and xanthine (Xan) in arterial blood samples obtained from 53 preterm infants during the first hour after birth. Cerebral ultrasound was performed within 48 h of birth and repeated at 5- or 6-day intervals until the age of 4 weeks.
Results: Grade I or II IVH was detected during the first 10 days of life in 11 of 53 patients (21%). Activin A, Hyp, and Xan concentrations and NRBC counts were higher in preterm newborns who subsequently developed IVH than in those who did not (P<0.0001, except P=0.019 for Xan). Neonatal activin A was correlated (P<0.0001) with Hyp (r=0.95), Xan (r=0.90), and NRBC count (r=0.90) in newborns without later IVH and in those who developed IVH (Hyp, r=0.89, P=0.0002; Xan, r=0.95, P<0.0001; NRBC count, r=0.90, P=0.0002). At a cutoff of 0.8 microg/L activin A, the sensitivity and specificity were 100% [11 of 11; 95% confidence interval (CI), 71%-100%] and 93% (39 of 42; 95% CI, 81%-98%), and positive and negative predictive values were 79% (95% CI, 61%-100%) and 0% (95% CI, 0%-2%), respectively. The area under the ROC curve was 0.98.
Conclusions: Activin A concentrations at birth are increased in preterm newborns who later develop IVH and may be useful for early identification of infants with hypoxic-ischemic brain insults who are at high risk for IVH.